INDEX

Explanations

attends to tokens related to relative pronouns or demonstrative pronouns from corresponding nouns or phrases

oai_attention-head · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 10-gemmascope-att-16k

Configuration

google/gemma-scope-2b-pt-att/layer_10/width_16k/average_l0_70

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.10.attn.hook_z

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Head Attr Weights

0:0.14

1:0.16

2:0.19

3:0.06

4:0.04

5:0.03

6:0.04

7:0.31

Negative Logits

DllImport

-0.31

jątk

-0.28

 quam

-0.27

 }}$.

-0.26

 where

-0.26

ovka

-0.26

))$.

-0.26

']").

-0.26

дайте

-0.24

льзя

-0.24

POSITIVE LOGITS

mybatisplus

0.41

:✨

0.39

HtmlAttribute

0.38

ArrowToggle

0.37

 Roskov

0.36

 Coton

0.33

 Cowan

0.32

 SwitchCompat

0.32

Джерела

0.31

 désolés

0.31

Activations Density 0.377%

attends to tokens related to relative pronouns or demonstrative pronouns from corresponding nouns or phrases

No Comments

No Known Activations